In square K9 of the Juukan 2 rock shelter, spit thirty-nine, nearly two metres below the modern surface, a fragment of tooth turned up in a four-millimetre sieve. It was the broken anterior half of a lower premolar, its root curved backward off the axis of the crown, its tip worn in the pattern of an animal that had already reached self-feeding age. Optically stimulated luminescence dating of the sand around it puts its age at somewhere between roughly 42,500 and 47,100 years, which makes it the oldest confirmed evidence that Tasmanian devils, Sarcophilus harrisii, ever lived in Western Australia’s Pilbara region. It is also, within the margin of the dating, about as old as people themselves are known to have been living in that rock shelter.
That fact runs against most people’s sense of where the species belongs. Tasmanian devils survive today only on the island whose name they carry, and even in the deeper fossil record, where their range once stretched across the mainland, the map has always had a hole in its northwest corner. More than 120 Quaternary sites across New South Wales, South Australia, Queensland, the Northern Territory, and southern, central, and tropical northern Western Australia have produced devil bones. The Pilbara itself, one of the hottest and driest stretches of the continent, had produced none, despite petroglyphs at Nganjarli, on the coast in Murujuga, that include figures researchers have read as devil-like among the thylacines and quolls carved into the same rock. A single tooth fragment from Barrow Island had been floated as a possible devil specimen in 2017, but nobody had confirmed it. Two specimens from Juukan Gorge, the Pleistocene premolar and a second, younger jaw fragment, close that gap.
The younger specimen is a piece of right lower jaw, recovered in a different part of the shelter, still holding its first and second molars. Charcoal from a hearth just above it and another just below bracket its age to somewhere between about 6,900 and 7,300 years, well into the Holocene. Both molars are ground down, their forward cusp worn away entirely, wear consistent with a long adult life rather than any damage after burial. Telling a devil’s tooth from a dingo’s, a quoll’s, or a thylacine’s is not obvious from a fragment like this. Devils and their dasyuromorphian relatives carry two similarly sized, multi-cusped molars in a row, where a dingo’s lower jaw carries one long shearing blade of a tooth behind a small, reduced second molar. Within that group, species of Sarcophilus can be told apart from thylacines and quolls by details as fine as which side of a molar cusp sits toward the cheek rather than the tongue, and, in the premolars, by a backward-curved root that no other Australian carnivore shares. The premolar was also about a quarter smaller than a typical adult devil’s, a discrepancy the research team1 attributes to enamel flaking, possibly from heat, rather than to a different or smaller species. Working from these features, and from a body-mass estimate of roughly seven to fourteen kilograms drawn from molar length, the team assigned both fragments to the living species rather than to either of the larger, extinct Pleistocene devils, Sarcophilus laniarius or Sarcophilus moornaensis.
